My Research - Detecting Cancer with Positron Emission Tomography

Hi Readers,

I periodically have these blogging sprees like today. It's too bad they don't come more often than they do. Right now, many of you probably know that I am at Stanford University, en route to getting a Ph.D in electrical engineering. However, I don't think I've ever talked about what kind of research I do there.

Like the title suggests, I work in a group  that focuses primarily on a medical imaging method called Positron Emission Tomography (PET). It is a technique commonly used for "functional imaging", which means that you are trying to image a process in the body, instead of the structure of the body. Most conventional imaging, such as CT scans (shooting x-rays through the body), MRIs (using magnetic fields to induce proton rotation in the body), and ultrasounds (bouncing sound waves through the body), use structural imaging to accomplish their task.

Here's a simple explanation of how PET works for cancer imaging. Cancer tumors are simply clumps of cells whose only orders are to grow as fast as they can, and forget about what they were originally supposed to do. To grow, they must have as much sugar as possible, and so they gobble up all the sugars in the body to maximize their growth rate. Therefore, if we can attach some kind of tracking device to sugar, we should be able to see where the tumors are by seeing where the sugars go. This is exactly how PET works.

In the case of PET, the "tracking device" is actually a positron that the specialized sugar (called FDG) releases in the body. When the positron is released, it will hit an electron in the body; since one is matter and one is antimatter, they will disappear into pure energy (called annhilation). In this case, the pure energy is released as two photons, or light particles, which shoot out of the location of annihilation in the complete opposite direction of each other. As these two leave the body, they hit detectors that surround the body, which records down their location. After that, you simply draw a line between these 2 locations (like connect the dot), and now you have a line which shows the possible places the tumor is located. These lines go in every random direction and share only one common trait, which is that they must intersect the tumor somewhere. Therefore, if you draw a million lines, the place where they all cross each other is the place where the tumor is located.

The project that I am working on involves the construction of a PET machine that can be used specifically for breast cancer imaging. There are five members on the project team. I am the person in charge of designing and integrating hardware to monitor the operation of the machine. This involves prototyping different configurations of the signal lines on the machine, sensors for temperature and humidity, as well as the detector-by-detector monitoring. Of course, it involves more than that, and probably has much more in common with an industry job than a university Ph.D, but it is a fulfilling and exciting role on a team of brilliant people.

Till next post!
FCDH